Study On The Forecasting And Optimization Of Short-Term Traffic Flow Dynamics Modeling Based On Grey Theory

With the rapid development of world economy and the accelerating of urbanization process, traffic problem has become a dilemma all over the world. Intelligent transportation system (ITS) is considered to be one of the effective methods to alleviate urban traffic problems, while road short-term traffic flow analysis is the core of ITS. Therefore, research on the effective modeling theory and method for short-term traffic flow is of great theoretical significance and practical value. The traffic flow theory is to research the law of road traffic by using mathematics and mechanics theory. Shot-term traffic flow system is largely similar to hydrodynamics system and has the basic characteristics of dynamics system. While this system is with highly uncertainty and it is difficult for people to grasp the influence scope and mechanism of the factors in the system, which belongs to the poor information; If one hour is treated as one period and the recording interval is taken as5minutes, then only12sets of data are collected, which belongs to the small sample data. Therefore, using grey theory which is dealing with small sample and poor information system to research the short-term traffic flow system with dynamics characteristics is reasonable and feasible. The main content of this paper is to study on the forecasting an optimization of short-term traffic flow dynamics modeling based on grey theory.In order to understand the mechanism of traffic flow, this paper analyzes the three basic dynamical properties of traffic flow system:continuity, viscosity and compressibility. It is shown that vehicle behavior will impact the traffic condition and the influence will be transmitted to the whole system. As the viscous dynamics characteristic of traffic flow, the impact need some time passing, so it showed that traffic flow system is of delay effect. As the pressure and viscous force makes the traffic flow system in complex and presenting nonlinear phenomena.The grey traffic flow dynamic model is constructed based on the grey characteristic of traffic flow and the classical grey theory, and its modeling mechanism is researched. Firstly, using the idea of matrix analysis, the modeling data matrix B is decomposed into three matrixes which respectively relate to accumulated generating, background value and original data processing. Based on the results, this paper proposes the generalization form of the classical grey model. Then, we study the influence of model parameters and predictive value by the transformation of the three matrixes in generalization model. Finally, we optimized the extended model.According to the three different characteristics of real traffic flow series, we construct the corresponding traffic flow dynamics model. For traffic flow series nonlinear characteristics, we construct grey nonlinear traffic flow dynamic model, and analyze the characteristics and trend of model’s curve; For traffic flow series of delay effect, we construct grey delay traffic flow dynamics model, determine the delay time by using the three factors relationship in traffic system and analyze the characteristic of the model; For multiple-section traffic flow of nonlinear and delay effect, we construct grey multiple factor delay and nonlinear traffic flow dynamic model, select the main factors by using grey relational analysis and introduce nonlinear factor and delay factor in MGM(1,N), and search the nonlinear factor and delay factor by using PSO.In order to understand the traffic flow dynamic system better, we research on the decision, optimization and control problems which are related to the system. To solve the problem of decision-making problem in road traffic system users, a grey relation analysis method based on two-tuple linguistic consistency is proposed and a unifier method to switch all the decision information to two-tuple linguistic information is provided; According to the mutual influence of participants in traffic system, we research grey bi-level multi-objective linear programming problem and analyze the solving method for the drifting grey bi-level multi-objective linear programming problem detailedly; For the problem of single control mode of intelligent signal, a method for intelligent signal dynamic optimization is proposed on the basis of the trend and the change of trend.